In a gene expression array study comparing signatures of chronic liver diseases with hepatocellular carcinoma (HCC), we found a molecular signature that separates patients for their risk of developing advanced disease. Epithelial cell adhesion molecule (EpCAM) was identified as the lead gene and its silencing resulted in growth suppression of HCC cells. EpCAM could significantly differentiate HCC into two subtypes. HCC could be further stratified into four distinct subtypes with the additional assessment of alpha-fetoprotein (AFP) status. These four subtypes were associated with HCC prognostic outcome and cells double positive for EpCAM and AFP had the worst prognosis. Furthermore, these subtypes resembled certain stages of liver lineages and EpCAM/AFP-positive cells displayed a distinct molecular signature with features of hepatic stem/progenitor cells. Moreover, these cells, characterizing a poor prognostic HCC subtype, were capable of initiating highly invasive HCC in in-vitro and in-vivo models. Thus, EpCAM and AFP are useful diagnostic markers for HCC which can be used as a convenient classification system for prognosis. Furthermore, EpCAM and AFP may act as downstream molecules to maintain HCC stemness and as markers for HCC initiating cells. We have also recently explored whether integrative genomic profiling of a well-defined HCC subset of extreme EpCAM+ AFP+ could uncover survival-related driver genes in HCC. We found that YY1-associated protein 1 (YY1AP1) is a critical oncoprotein specifically activated in EpCAM+AFP+ HCC. YY1AP1 silencing eliminates oncogene addiction by altering the chromatin landscape while YY1AP1 expression promotes HCC proliferation and is required for the maintenance of stem cell features. Thus YY1AP1 may serve as a key molecular target for EpCAM+ AFP+ HCC subtype. We also investigated the mechanism by which EpCAM is elevated in HCC subtypes with stem/progenitor cell features. We found that the activation of wnt-beta-catenin pathway regulates EpCAM expression. In fact, EpCAM is a biosensor for wnt-beta-catenin signaling and is transcriptionally up-regulated by this pathway through direct Tcf binding element interactions. Inhibition of HCC cell growth could be achieved through blockade of EpCAM/wnt-beta-catenin signaling in EpCAM-positive HCC cells. We propose that EpCAM/wnt-beta-catenin signaling functions to maintain HCC stem cell growth and that EpCAM expression-based classification of HCC could be useful in clinical settings to stratify HCC patients who may benefit from beta-catenin/EpCAM adjuvant therapies. We recently developed a high-throughput screening assay to identify inhibitors of EpCAM-dependent growth of HCC cells. EpCAM(+) and EpCAM(-) HCC cell lines were assessed for differential sensitivity to Wnt/beta-catenin pathway inhibitors. Libraries comprising 22 668 pure compounds and 107 741 crude or partially purified natural product extracts were tested, and 12 pure compounds and 67 natural product extracts were identified for further study. Three active compounds and the positive control were further characterized by effects on EpCAM expression. Treatment of EpCAM(+) Hep3B cells resulted in loss of EpCAM expression by flow cytometry. The identification of compounds with a variety of possible molecular targets suggests a likelihood of multiple mechanisms for modulation of EpCAM-dependent cell growth. Recently, we found that an FDA-approved psychiatric drug, pimozide (PMZ), has anti-cancer properties in HCC cell lines that express EpCAM. We demonstrate that PMZ effectively inhibits cell growth of HCC cells by disrupting the wnt/beta-catenin signaling pathway and reducing EpCAM expression. Thus, PMZ may be a useful molecular entity that could be repurposed as an anti-cancer therapy for treatment of HCC. A global microRNA microarray approach was used to explore whether certain microRNAs were associated with HCC stem cells. We found that microRNA-181 family members were up-regulated in HCC stem cells. Inhibition of microRNA-181 led to a reduction in number and tumor initiating activity of HCC stem cells while addition of microRNA-181 led to an enrichment of this cell type. We also showed that microRNA-181 could directly target transcriptional regulators of differentiation in the liver and an inhibitor of wnt-beta-catenin signaling. In addition, we have recently shown that Wnt/beta-catenin signaling transcriptionally activates microRNA-181s in HCC. These results suggest a novel regulatory link between microRNA-181 family members, Wnt/beta catenin signaling and liver cancer stem cells and implies that molecular targeting of microRNA-181 or Wnt/beta-catenin signaling may eradicate hepatocellular carcinoma. We have also recently explored whether specific microRNAs exist in hepatic cancer stem cells (CSCs) that are not expressed in normal hepatic stem cells by assessing the microRNA transcriptome of HCC specimens by small RNA deep sequencing. We found that miR-150, miR-155, and miR-223 were preferentially highly expressed in EpCAM+ HCC cells and their gene surrogates were associate with patient prognosis. Furthermore, suppressing miR-155 reduced EpCAM+ HCC cells, HCC tumorigenicity and shortened overall survival and time to recurrence of HCC patients. Thus, miR-155 might serve as a molecular target to eradicate the EpCAM+ CSC population in human HCCs. We have recently assessed molecular signatures related to HCC stemness and outcome in intrahepatic cholangiocarcinoma (ICC). Using Affymetrix mRNA and Nanostring microRNA microarrays, Asian ICC cases could be segregated into two main subgroups, one of which shared gene expression signatures with previously identified HCC with stem cell gene expression traits. Integrative analyses of the ICC-specific mRNA and microRNA expression profiles revealed that miR-200c and epithelial-mesenchymal transition (EMT) signaling was preferentially activated in ICC with stem cell gene expression traits. Inactivation of miR-200c resulted in an induction of EMT while its activation reduced EMT and cell migration and invasion. NCAM1, a hepatic stem/progenitor cell marker, was a direct target of miR-200c. Our results indicate that ICC and HCC share common stem-like molecular characteristics and poor prognosis and that specific EMT components may be critical biomarkers and clinically relevant therapeutic targets for an aggressive form of stem cell-like ICC. Pre-clinical research to delineate molecular mechanisms that drive cancer growth and progression is usually carried out in two-dimensional (2D) cell cultures. They are efficient and reliable but lack the appropriate cell-cell contact environment typically observed in vivo. Thus the 3D organotypic model provides an important alternative to both 2D culture and in vivo animal model systems. Recently, we demonstrated that HCC cells in porous alginate scaffolds can generate organoid-like spheroids that mimic numerous features of glandular epithelium in vivo, such as acinar morphogenesis and apical expression patterns of EpCAM. We show that the activation of Wnt/beta-catenin signaling, a pathway for maintaining HCC stemness, is required for EpCAM+ HCC spheroid formation and the maintenance of the acinous structure. We also demonstrated that EpCAM+ HCC cells cultured as spheroids are more sensitive to TGF-induced epithelial-mesenchymal transition with highly tumorigenic and metastatic potential in vivo compared to conventional 2D culture. HCC cells in EpCAM+ spheroids are more resistant to chemotherapeutic agents than 2D-cultured cells. The alginate scaffold-based organotypic culture system is a promising, reliable, and easy system that can be configured into a high-throughput fashion for the identification of critical signaling pathways and screening of molecular drug targets specific for HCC.